{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:AUAKEHVYAO55R4KELVSPUKQPSH","short_pith_number":"pith:AUAKEHVY","schema_version":"1.0","canonical_sha256":"0500a21eb803bbd8f1445d64fa2a0f91eefe31fd4216df5b815c2a64119c65e9","source":{"kind":"arxiv","id":"2505.17941","version":1},"attestation_state":"computed","paper":{"title":"VeriThinker: Learning to Verify Makes Reasoning Model Efficient","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Gongfan Fang, Ruonan Yu, Xinchao Wang, Xinyin Ma, Zigeng Chen","submitted_at":"2025-05-23T14:17:56Z","abstract_excerpt":"Large Reasoning Models (LRMs) excel at complex tasks using Chain-of-Thought (CoT) reasoning. However, their tendency to overthinking leads to unnecessarily lengthy reasoning chains, dramatically increasing inference costs. To mitigate this issue, we introduce VeriThinker, a novel approach for CoT compression. Unlike conventional methods that fine-tune LRMs directly on the original reasoning task using synthetic concise CoT data, we innovatively fine-tune the model solely through an auxiliary verification task. By training LRMs to accurately verify the correctness of CoT solutions, the LRMs inh"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2505.17941","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-05-23T14:17:56Z","cross_cats_sorted":[],"title_canon_sha256":"fcfb294bb0095ad8d4a15d01fa9be4726dd4de48480c678905c99c30bbba1676","abstract_canon_sha256":"c2b1d524007d9581ece65cb0b9228dc79fec64adaf49024e2d6be52e05e4e958"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:08:34.062881Z","signature_b64":"eVXh0NouGtzcoi41crGNLSf4NMPRvN1LJbda5ER39+QqrQ91y9G2PFUrbYBIzrVPKZ5ImqAjRcrxz2vO8RWtBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0500a21eb803bbd8f1445d64fa2a0f91eefe31fd4216df5b815c2a64119c65e9","last_reissued_at":"2026-07-05T11:08:34.062489Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:08:34.062489Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"VeriThinker: Learning to Verify Makes Reasoning Model Efficient","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Gongfan Fang, Ruonan Yu, Xinchao Wang, Xinyin Ma, Zigeng Chen","submitted_at":"2025-05-23T14:17:56Z","abstract_excerpt":"Large Reasoning Models (LRMs) excel at complex tasks using Chain-of-Thought (CoT) reasoning. However, their tendency to overthinking leads to unnecessarily lengthy reasoning chains, dramatically increasing inference costs. To mitigate this issue, we introduce VeriThinker, a novel approach for CoT compression. Unlike conventional methods that fine-tune LRMs directly on the original reasoning task using synthetic concise CoT data, we innovatively fine-tune the model solely through an auxiliary verification task. By training LRMs to accurately verify the correctness of CoT solutions, the LRMs inh"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.17941","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2505.17941/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2505.17941","created_at":"2026-07-05T11:08:34.062543+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.17941v1","created_at":"2026-07-05T11:08:34.062543+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.17941","created_at":"2026-07-05T11:08:34.062543+00:00"},{"alias_kind":"pith_short_12","alias_value":"AUAKEHVYAO55","created_at":"2026-07-05T11:08:34.062543+00:00"},{"alias_kind":"pith_short_16","alias_value":"AUAKEHVYAO55R4KE","created_at":"2026-07-05T11:08:34.062543+00:00"},{"alias_kind":"pith_short_8","alias_value":"AUAKEHVY","created_at":"2026-07-05T11:08:34.062543+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2512.19728","citing_title":"Hard Negative Sample-Augmented DPO Post-Training for Small Language Models","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2503.16419","citing_title":"Stop Overthinking: A Survey on Efficient Reasoning for Large Language Models","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04120","citing_title":"Shorter, but Still Trustworthy? An Empirical Study of Chain-of-Thought Compression","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11625","citing_title":"Nice Fold or Hero Call: Learning Budget-Efficient Thinking for Adaptive Reasoning","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06165","citing_title":"Post Reasoning: Improving the Performance of Non-Thinking Models at No Cost","ref_index":241,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/AUAKEHVYAO55R4KELVSPUKQPSH","json":"https://pith.science/pith/AUAKEHVYAO55R4KELVSPUKQPSH.json","graph_json":"https://pith.science/api/pith-number/AUAKEHVYAO55R4KELVSPUKQPSH/graph.json","events_json":"https://pith.science/api/pith-number/AUAKEHVYAO55R4KELVSPUKQPSH/events.json","paper":"https://pith.science/paper/AUAKEHVY"},"agent_actions":{"view_html":"https://pith.science/pith/AUAKEHVYAO55R4KELVSPUKQPSH","download_json":"https://pith.science/pith/AUAKEHVYAO55R4KELVSPUKQPSH.json","view_paper":"https://pith.science/paper/AUAKEHVY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.17941&json=true","fetch_graph":"https://pith.science/api/pith-number/AUAKEHVYAO55R4KELVSPUKQPSH/graph.json","fetch_events":"https://pith.science/api/pith-number/AUAKEHVYAO55R4KELVSPUKQPSH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AUAKEHVYAO55R4KELVSPUKQPSH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AUAKEHVYAO55R4KELVSPUKQPSH/action/storage_attestation","attest_author":"https://pith.science/pith/AUAKEHVYAO55R4KELVSPUKQPSH/action/author_attestation","sign_citation":"https://pith.science/pith/AUAKEHVYAO55R4KELVSPUKQPSH/action/citation_signature","submit_replication":"https://pith.science/pith/AUAKEHVYAO55R4KELVSPUKQPSH/action/replication_record"}},"created_at":"2026-07-05T11:08:34.062543+00:00","updated_at":"2026-07-05T11:08:34.062543+00:00"}